Estudos anteriores têm relatado um decremento na capacidade de produção de força com a idade. Mudanças estruturais e funcionais nos músculos esqueléticos tal como a sarcopenia (perda de massa muscular), diminuição da ativação das unidades motoras, co-contração aumentada da musculatura antagonista, diminuição da tensão da fibra muscular e rigidez do tendão são referidos para explicar a funcionalidade motora reduzida em indivíduos idosos. Este estudo objetivou investigar as mudanças ligadas à idade na produção de torque voluntário isométrico e isocinético (TVM), como também as respostas eletromiográficas (EMG). Hipotetizamos, como conseqüências do envelhecimento: a) uma redução da capacidade de produção de TVM; b) deslocamento da curva torque-ângulo (para a esquerda) na direção de menores comprimentos musculares; c) deslocamento da curva torque-velocidade (para a esquerda) na direção de menores velocidades; d) que uma menor ativação muscular seria observada. Quarenta indivíduos saudáveis, do gênero masculino, divididos igual e randomizadamente em quatro grupos (15-30; 31-45; 46-60 e 61-75 anos), participaram neste estudo como sujeitos, aos quais foram solicitados produzir torques máximos de extensão de joelho em um dinamômetro isocinético (Byodex®) em diferentes ângulos (15º; 30º;45º; 60º; 75º; 90º e 105º) e velocidades (60º/s; 120º/s; 180º/s; 240º/s; 300º/s e 360º/s) Sinais da EMG dos músculos da coxa, como vasto lateral, reto femoral e vasto medial foram simultaneamente registrados. O nível de atividade física foi avaliado pelo IPAQforma curta (International questionnaire of physical activity ). Valores de TVM, valores RMS do sinal EMG e nível de atividade física foram usados como medidas dependentes. Os resultados demonstraram uma diminuição no TVM com a idade. Particularmente, menores valores de TVM foram encontrados durante as maiores velocidades em condições isocinéticas. Nenhuma mudança foi observada nas curvas torque-ângulo e torque-velocidade e nenhuma mudança ligada à idade foram encontradas nas respostas da EMG. Concluímos que a idade afeta a produção de TVM, particularmente em maiores velocidades. Além disso, os resultados permitemnos apontar que os altos níveis de atividade física e demandas funcionais possam explicar a similaridade encontrada nas respostas EMG entre os diferentes grupos. / Previous studies have reported a decrease in force production ability with age. Structural and functional changes on the skeletal muscles such as sarcopenia (loss of muscle mass), decreased ability of motor unit firing, increased co-activation of antagonist muscle, decreased muscle fiber tension and tendon’s rigidity are taken into account for the reduced motor functionality in elderly individuals. This study aimed to investigate age-related changes of maximum voluntary isometric and isokinetic torque production (MVT), as well as electromyography (EMG) responses. We hypothesized that as consequence of aging: a) a reduced ability of MVT production; b) a torque/angle curve displacement (left side) towards the lowers muscle lengths; c) a torque/velocity curve displacement (left side) towards smaller velocities; d) and decreased muscle activation will be observed. Forty healthy male individuals, evenly and randomly divided in four age groups (15–30; 31– 35; 45– 60; 60–75 years old), participated in this study as subjects. The subjects were asked to produce maximum knee extension torques on the isokinetic dynamometer (Cybex ®) in different angles (15º; 30º; 45º; 60º; 75º; 90º and 105º) and velocities (60º/s; 120º/s; 180º/s; 240º/s; 300º/s and 360º/s) EMG signals from the upper leg muscles such as, vatsus lateralis, rectus femoris, vastus medialis, were simultaneously recorded. The level of physical activity was evaluated by the QIAF–short form (International questionnaire of physical activity). MVT values, RMS values from EMG signals and level of physical activity were used as dependent measures. The results showed a decrease in MVT with age. Particularly, lower values of MVT were found during higher velocities under isokinetic condition. No changes were observed for torque/angle and torque/velocity curves and no age-related changes were found on the EMG responses. We conclude that aging affects MVT production, particularly during high velocities. In addition, the results allowed us to point out that higher levels of physical activity and functional demand, showed by the older groups, could explain the similarities found on EMG responses among the different age groups.

No description available.

Context: Knee injury is a common problem for runners. Knee pain is a common symptom in knee injury and is associated with alterations in knee and hip muscle activation and hip joint angles. Relieving pain through intervention may help to restore neuromuscular function. Objective: To examine the effects of ice and transcutaneous electrical nerve stimulation (TENS) combination treatment on perceived knee pain, hip frontal plane angle, and muscle activation during running in individuals with experimental knee pain (EKP). Design: Crossover. Setting: Laboratory. Subjects: 19 participants (11 males and 8 females, 23.2 ± 1.9 y, 176 ± 11.6 cm, 71.5 ± 16.9 kg; right leg dominant). Interventions: Hypertonic saline was infused into the infrapatellar fat pad for 74 minutes (total 11.1 mL). Subjects underwent 2 treatment conditions (sham; ice/TENS combination). Measurements were recorded during running at 4 time points (preinfusion, postinfusion, posttreatment, and postinterval). Main Outcome Measures: Perceived knee pain on a 100-mm visual analog scale (VAS), knee and hip muscle peak electromyography (EMG) amplitude, and hip adduction angles. Results: Hypertonic saline infusion increased perceived anterior knee pain in all participants. The average of peak perceived knee pain was 28 mm on a 100-mm VAS in EKP application. While the increased perceived knee pain level stayed consistent across time in the sham session, ice/TENS combination treatment significantly reduced perceived knee pain by 35% at 6 minutes after the treatment start (p = 0.049), and the reduced knee pain lasted for 22 minutes (p > 0.05). Peak EMG amplitude of the gluteus medius was decreased by 13.5% and 14.3% (p = 0.023; p = 0.013) during running after EKP in sham and treatment sessions, respectively. However, the peak EMG amplitude was not restored to pain-free level during running after the treatment (p = 0.026). No other muscles changed their peak EMG amplitude due to EKP or treatment. Hip adduction angles during running were also not altered by EKP or treatment (p > 0.3) in both sham and treatment sessions. Conclusions: EKP increased perceived knee pain and decreased peak muscle activation of the gluteus medius during running. Ice/TENS combination treatment reduced perceived knee pain quickly, but did not restore neuromechanics during running.

No description available.

La stimulation électrique fonctionnelle (SEF) est une des techniques utilisées pour la rééducation ou la suppléance fonctionnelle de déficiences motrices. Un stimulateur génère des impulsions électriques qui induisent des contractions des muscles paralysés, au travers des unités motrices toujours intactes. Aujourd'hui, les systèmes fonctionnant en boucle ouverte sont majoritairement utilisés. Ils permettent aux chercheurs d'évaluer hors-ligne de la SEF. Cependant, leur réglage reste basé sur un cycle essai-erreur ce qui est loin de simplifier la mise en œuvre de systèmes en boucle fermée. Dans cette thèse, nous proposons une méthode de prédiction en temps-réel du couple en fonction de l'EMG conduisant à une nouvelle modalité de contrôle de l'activation musculaire sous SEF. L'EMG évoquée (eEMG) donne une image de l'effet de la SEF sur l'activité musculaire et est ainsi impliquée à la fois dans l'estimation du couple en temps réel et le contrôle. Le couple articulaire peut être estimé par l'eEMG en utilisant un filtre de Kalman et un modèle NARX - RNN. Le facteur d'oubli du filtre de Kalman doit être soigneuse-ment choisi de même que les réglages du schéma de calcul. C'est une limitation en particulier quand lorsque qu'il n'y a aucune connaissance a priori sur la force générée par le sujet sous SEF. La méthode proposée NARX-RNN n'a pas ce défaut et offre de meilleures performances que le filtre de Kalman. L'estimation du couple basée eEMG est donc utilisée hors-ligne et en-ligne en temps réel. Les performances comparées des algorithmes ont été effectues sur sujets sains et sujets blessés médullaires. Par ailleurs, le système temps-réel de contrôle de l'activation musculaire basé EMG a été développé sur la technologie sans fil Vivaltis. Enfin, afin de proposer un contrôle de plusieurs muscles, le concept de synergie a été utilisée pour estimer les activations musculaires cibles à partir d'un couple articulaire désiré. Les niveaux de synergie moyenne ont été utilisés pour valider l'extraction d'activation sur le sujet non inclus dans le calcul de la moyenne. L'erreur d'estimation est de 9.3% sur l'ensemble des sujets. Ces résultats vont dans le sens d'un contrôle d'une neuroprothèse basé synergie. En effet la combinaison des eux contributions de la thèse ouvre des perspectives nouvelles de modalité de contrôle de la SEF. / Functional electrical stimulation (FES) is one of existing rehabilitationtechniques to restore lost motor functions for motor-impaired subjects. Thestimulator generates electrical pulses to drive artificial contractions of theparalysed muscles, through activating intact motor units. Currently open-loopFES system is the most frequently used. The data acquired from the open-loop FESwould help researchers to make off-line analysis for evaluating performance ofFES systems. However, it should go through a trial and error manner, which isfar from facilitating a implementation of real-time closed-loop FES system.In this thesis, we propose and develop a method for real-time EMG-feedback torqueprediction and muscle activation control toward new modality in FES.The evoked electromyography (eEMG) which can reflect electrical muscleactivities under FES, is involved in both offline and real-time FES-inducedtorque estimation and muscle control systems. FES-induced joint torque can beestimated/predicted with eEMG by employing both Kalman filter and NonlinearAuto-Regressive with Exogenous (NARX) type recurrent neural network (RNN). Theforgetting factor of Kalman filter should be properly selected in advance andalso with proper computational settings. It is a limitation for some casesespecially when we do not have prior knowledge of new subject regarding expectedmuscle response intensity induced by FES. The proposed NARX-RNN does not sufferfrom such computational setting problems and also shows better estimation/prediction performances than that of Kalman filter.Evoked EMG based torque estimator is exploited from off-line situation toonline real-time system. Recursive Kalman filter and NARX-RNN are implementedfor real-time torque estimation/prediction with evoked EMG. The performance wasverified both in able-bodied and spinal cord injured subjects. Furthermore, real-time EMG-feedback muscle activation control in FES system is developed togetherwith wireless Vivaltis stimulator for specifying directly muscle activationinstead of conventionally specifying stimulation pattern.Toward natural multiple muscles control with multi-channel FES, muscle synergyconcept was introduced for inverse estimation of muscle activations from desiredjoint moment. The averaged synergy ratio was applied for muscle activationestimation with leave-one-out cross validation manner, which resulted in 9.3%estimation error over all the subjects. This result supports the common musclesynergy-based neuroprosthetics control concept. By combining this inverse estimation of muscle activations together with real-time EMG-feedback muscle activation control, it would open a new modality toward muscle synergy-basedmulti-muscle activation control in FES.

The very short-term resistance training (VST) model, utilizing only 2-3 training sessions, has been used to examine early phase skeletal muscle, neural, and performance adaptations. The VST model has previously been used to examine these early phase adaptations in bilateral and unilateral, isometric, isokinetic, and dynamic muscle actions in the limbs of the upper- and lower-body. The bilateral deficit (BLD) is a phenomenon in which the sum of the forces produced unilaterally is greater than the force produced bilaterally during maximal contraction of the limbs. The appearance of a bilateral deficit has been be related to various factors; including training status and mode of training (bilateral versus reciprocal muscle actions). No previous study, however, has examined the effects of VST on the BLD. The VST model has potential implications for examining acute changes in strength and neuromuscular responses of the trained muscles. These adaptations, however, may be specific to unilateral or bilateral training. Therefore, the purposes of this study were to: 1) examine one repetition maximum (1RM) strength and neuromuscular responses (EMG AMP, EMG MPF, MMG AMP, MMG MPF) during the measurement of bilateral and unilateral leg extension exercise before and after dynamic constant external resistance (DCER) VST; 2) examine the magnitude of the BLD; 3) examine the effect of bilateral versus unilateral training on the BLD; and 4) use the neuromuscular responses measured bilaterally and unilaterally to infer about the motor unit activation strategies that may underlie the BLD and changes in 1RM strength. Twenty-four (14 males, 10 females) subjects (mean ± SD age: 23.0 ± 3.2 yr; height: 174.7 ± 8.5 cm; body mass: 75.4 ± 14.1 kg) with no resistance training experience within the last three months were randomly assigned to either the bilateral (BL) training group or the unilateral (UL) training group. The subjects completed a total of seven visits, consisting of a familiarization, pre-test visit, three training visits, and one post-test visit. The pre-test visit was used to record the subject’s electromyographic (EMG) and mechanomyographic (MMG) responses from the right and left vastus lateralis (VL) during bilateral and unilateral seated maximum isometric voluntary contractions (MVIC) and 1RM. Visits four through six were the training sessions, with each subject preforming 5 sets of 6 repetitions utilizing 65% of the 1RM for resistance where the BL group trained both limbs (right and left) at the same time and the UL group trained both limbs separately. Visit seven was the post-test and the same testing procedures as the pre-test visit were followed. Statistical analyses consisted of four-way and three-way mixed model ANOVAs, with follow up three-, two- and one-way repeated measures and/or mixed model ANOVAs, Bonferroni corrected paired, and independent samples t-tests when appropriate. An alpha level of P ≤ 0.05 was considered statistically significant for all ANOVAs. The BL group demonstrated a significant increase (p = 0.006; 6.8%) in BL1RM pre- to post-test, but no change in unilateral summed (US1RM = right + left limb; p = 0.726) 1RM strength. The UL group demonstrated an 8.7% increase in BL strength collapsed across testing mode (BL1RM and US1RM) (p = 0.0001) and UL strength (p = 0.0001) collapsed across limb (UL left + UL right/2) from pre- to post-test. The BL group had a significant (p = 0.001) increase in the BI (indicating a decrease in the BLD) from pre- to post-test, but there was no significant change for the UL group. The BL group demonstrated a significant (p = 0.029) decrease in the EMG mean power frequency (MPF) measurement pre- to post-test, however the UL group showed no change. The unilateral movement, collapsed across limbs (unilateral left and unilateral right) also showed a significant (p = 0.022) decrease in the MMG MPF measurement pre- to post-test, whereas the BL movement showed no change. These findings indicated that BL and UL DCER training increased strength after 3 training sessions. The bilateral DCER training resulted in bilateral, but not unilateral strength increases and unilateral DCER training resulting in both bilateral and unilateral strength increases. However, bilateral training was the only mode of training that significantly decreased the BLD.

L'objectif de ce travail est de contribuer au développement de méthodes de choix de la meilleure base à partir d'une décomposition en paquets d'ondelettes pour la détection et la classification. Le cadre applicatif global est le choix de la meilleure base pour la détection d'événements dans le signal EMG utérin, utilisé pour la prévention des accouchements prématurés. Deux approches de modélisation sont utilisées pour mettre en évidence le contenu fréquentiel des événements. La première est fondée sur la décomposition discrète en ondelettes, la deuxième sur la décomposition en paquets d'ondelettes. La distance de Kullback Leibler est utilisée comme un critère du choix de la meilleure base pour la détection. La détection est effectuée sur les coefficients des paquets sélectionnés. Un décalage est généré différemment sur chaque paquet d'où la nécessité de redéfinir les vraies valeurs des instants de changement et d'appliquer une procédure de fusion pour avoir ensuite un seul instant de détection correspondant au signal original. Le choix des paquets les plus discriminants pour la classification est traité. Les événements détectés sont identifiés "physiologiquement" en utilisant les méthodes de K Plus Proches Voisins, la distance de Mahalanobis, les réseaux de neurones et les Machines à Vecteurs Support. Plus de 85% des événements ont été bien classifiés.

Cette recherche a été menée dans le contexte scientifique proposé par l'équipe BIOMED (Pr. H. RIX), équipe qui regroupe des chercheurs du laboratoire I3S et des médecins (F. RAYBAUD, G. SUISSE) , Praticiens Hospitaliers, ayant une formation scientifique complémentaire (DEA ou Doctorat SPI). Cette association permet d'aborder des problèmes réels posés par la Médecine ou la Physiologie concernant l'interprétation des signaux biomédicaux. Mon activité au sein de cette équipe se situe dans le domaine du Traitement du Signal pour ce qui concerne l'élaboration de nouveaux concepts ou outils et en amont du Génie Biologique et Médical pour les applications. Dans le même type d'activité, on peut citer les équipes françaises de Jean-Louis COATRIEUX (INSERM- Rennes 1), Catherine MARQUE (CNRS-UTC), Jacques DUCHENE (UTT), Paul RUBEL (INSERM- INSA Lyon). Les principaux domaines d'application actuels sont l'électrocardiographie (ECG), les signaux électriques cérébraux (EEG et PE) et l'électromyographie (EMG) de surface. Les signaux étudiés sont donc des transitoires bruités souvent répétitifs. Ce sont soit des signaux 1D soit des signaux vectoriels constitués d'enregistrements simultanés provenant de capteurs diversement positionnés ou d'enregistrements successifs d'un même capteur. Les problèmes généraux rencontrés sont l'estimation de signaux moyens associés le plus souvent à des phénomènes permanents ou stables au cours du temps, la caractérisation de la variabilité naturelle et enfin la mesure de variations significatives de paramètres ou plus généralement de forme, liées à un agent extérieur tel qu'une pathologie, un effort ou un médicament. Les approches théoriques relèvent de la détection, de l'estimation de modèles paramétriques, semi-paramétriques ou non paramétriques et de la classification, dans le cas non stationnaire.

<p>Human skeletal muscles possess thixotropic, i.e. history-dependent mechanical properties. This means that the degree of passive muscle stiffness and resting tension is dependent on the immediately preceding history of contractions and length changes. Athletes, for instance, reduce passive muscle stiffness by various types of ‘limbering-up’ procedures, whereas muscle stiffness gradually increases during inactivity.</p><p>Passive resistance of antagonistic muscles may significantly add to the total load during voluntary muscle contractions. This resistance may vary from one moment to another, depending on immediately preceding events. This research was conducted to determine whether history-dependent variations in passive muscular forces influence motor control of voluntary joint movements and steady maintenance of joint positions in healthy subjects. </p><p>In study I, the EMG signal revealed motor compensations for history-dependent variations in passive stiffness of the antagonists during slow voluntary wrist joint movements. Studies II and III demonstrated that the voluntary muscle activity required to maintain a certain wrist joint position was highly influenced by previous changes in forearm muscle length and contractions. Study IV showed that rapid voluntary movements varied in speed and onset time depending on the prevailing degree of muscle resistance, and in addition that the central nervous reaction time required to execute rapid movements was highly influenced by immediately preceding muscle-conditioning procedures.</p><p>History-dependent variations in passive muscular forces seem to be effectively compensated by the motor control system. Presumably, voluntary motor commands to the muscles are automatically adjusted in strength to history-dependent changes in passive muscular forces. Such adjustments occur within the central nervous system, which receives information about the mechanical state of the muscles. Several issues in connection with muscle thixotropy remain unaddressed. For instance, do alterations in the normal thixotropic mechanical behaviour of the muscles impose a particular problem in patients with certain neuromuscular diseases? </p>

Spinal cord injury (SCI) manifests signs of spasticity, plantar flexor (PF) hyperreflexia and ankle clonus, and deficits in motor function. In individuals with motor-incomplete SCI (MISCI), ankle clonus may limit independent walking function. Ankle clonus is attributed to enhanced soleus stretch reflex (SSR) excitability due to decreased supraspinal input and maladaptive reorganization of spinal reflex circuitry. We explored these questions: 1. What are the biomechanical, clinical, and neurophysiologic correlates of ankle clonus? 2. Does locomotor training improve ankle clonus and walking function? 3. Will operant conditioning-based interventions that increase tibialis anterior activation or decrease soleus reflex excitability improve ankle motor control and walking function? In Chapter 2 we compared Ankle Clonus Drop Test (Drop Test) measures with clinical and neurophysiologic measures. Drop Test measures were highly reliable and exhibited moderate to strong correlations with clinical and neurophysiologic measures. Analysis of EMG activity during clonus revealed a predominant pattern of antagonist coactivation. In Chapter 3 we investigated the effects of locomotor training on PF and quadriceps spasticity, and walking function. We assessed responsiveness of the PF reflex threshold angle, a Drop Test measure of PF spasticity. PF and quadriceps spasticity decreased after locomotor training and were moderately correlated with increased walking speed. The PF reflex threshold angle measure discriminated between individuals with and without clonus. In Chapter 4 we compared the effects of two operant-conditioning based interventions to, (1) increase TA EMG activation (TA↑) and (2) decrease SOL H-reflex amplitude during active dorsiflexion (SOL↓), on reflex modulation, ankle motor control, and walking function. Each intervention improved walking function; however, modulated the variables in unique ways. TA↑ improved deficits of strength and range of motion, and SOL↓ improved modulation of SSR and SOL/TA coactivation. In Chapter 5 we discussed implications of our conclusions: (1) Drop Test ankle clonus measures are valid, reliable, and responsive; (2) antagonist coactivation was predominant during ankle clonus; (3) in individuals with chronic MISCI, locomotor training decreased PF and quadriceps spasticity and improved walking function; and (4) an operant conditioning-based intervention to either increase TA strength or decrease SOL reflex excitability improved spinal reflex modulation and walking function.